96th Congress \ COMMITTEE PIMTSTT / COMMITTEE2d Session / COMMITTEE PRINT j pRINT QQ_IFG ̂

AN EVALUATION OF THE STRATEGICPETROLEUM RESERVE

A REPORTPREPARED A^T THE REQUEST OF THE

SUBCOMMITTEE ON ENERGY AND POWER

COMMITTEE ONINTERSTATE AND FOREIGN COMMERCE

UNITED STATESHOUSE OF REPRESENTATIVES

BY THE

CONGKESSIONAL BUDGET OFFICE

JUNE 1980

Printed fcr the use of the Committee on Interstate and ForeignCommerce, United States House of Representatives

U.S. GOVERNMENT PRINTING OFFICE

63-724 O WASHINGTON : 1980

COMMITTEE ON INTERSTATE AND FOREIGN COMMERCE

BARLEY O. STAGGERS,JOHN D. DINGELL, MichiganLIONEL VAN DEER LIN, CaliforniaJOHN M. MURPHY, New YorkDAVID E. SATTERFIELD III, VirginiaBOB ECKHARDT, TexasRICHARDSON PREYER, North CarolinaJAMES H. SCHEUER, New YorkRICHARD L. OTTINGER, New YorkHENRY A. WAXMAN, CaliforniaTIMOTHY E. WIRTH, ColoradoPHILIP R. SHARP, IndianaJAMES J. FLORIO, New JerseyANTHONY TOBY MOFFETT, ConnecticutJIM SANTINI, NevadaANDREW MAGUIRE, New JerseyEDWARD J. MARKEY, MassachusettsTHOMAS A. LUKEN, OhioDOUG WALGREN, PennsylvaniaALBERT GORE, JR., TepnesseeBARBARA A. MIKULSKI, MarylandRONALD M. MOTTL, OhioPHIL GRAMM, TexasAL SWIFT, WashingtonMICKEY LELAND, TexasRICHARD C. SHELBY, AlabamaROBERT T. MATSUI, California

, West Virginia, ChairmanJAMES T. BKOYHILL, North CarolinaSAMUEL L. DEVINE, OhioTIM LEE CARTER, KentuckyCLARENCE J. BROWN, OhioJAMES M. COLLINS, TexasNORMAN F. LENT. New YorkEDWARD R. MAD1OAN, IllinoisCARLOS J. MOORHEAD, CaliforniaMATTHEW J. RINALDO, New JerseyDAVE STOCKMAN, MichiganMARC L. MARKS, PennsylvaniaTOM CORCORAN, IllinoisGARY A. LEE, New YorkTOM LOEFFLER, TexasWILLIAM E. DANNEMEYER, California

KENNETH J. PAINTER, Chief Clerk and Staff DirectorELEANOR A. DINKINS, First Assistant Clerk

WM. MICHAEL KITZMILLER, Professional StaffRANDALL E. DAVIS, Associate Minority Counsel

SUBCOMMITTEE ON ENERGY AND POWER

JOHN D. DINGELL, Michigan, ChairmanCLARENCE J. BROWN. OhioCARLOS J. MOORHEAD, CaliforniaJAMES M. COLLINS, TexasDAVE STOCKMAN, MichiganTOM CORCORAN, IllinoisTOM LOEFFLER. TexasJAMES T. BROYHILL, North Carolina

(Ex Offlcio)

RICHARD L. OTTINGER, New YorkPHILIP R. SHARP, IndianaANTHONY TOBY MOFFETT. ConnecticutDAVID E. SATTERFIELD III, VirginiaTIMOTHY E. WIRTH. ColoradoEDWARD J. MARKEY, MassachusettsPHIL GRAMM. TexasAL SWIFT. WashingtonRICHARD C. SHELBY, AlabamaANDREW MAGUIRE. New JerseyALBERT GORE, JR.. TennesseeMICKEY LELAND, TexasHARLEY O. STAGGERS, West Virginia

(Ex Offlcio)FRANK M. POTTER, Jr., Staff Director

MICHAEL F. BARRETT, Jr., CounselROGER P. STAIGER, Jr., Research Analyst

(LI)

LETTER OF TRANSMITTALNINETY-SIXTH CONGRESS v ROOM 3302

HOUSE OFFICE BUILDING ANNEX NO. 2

JOHN D. MNOU. MICH, CHAIRMAN PHONE (202) 225-1030

>«H« 3. -OWN. OH* CONGRESS OF THE UNITED STATES

HOUSE OF REPRESENTATIVESSUBCOMMITTEE ON ENERGY AND POWER

OF THECOMMITTEE ON INTERSTATE AND FOREIGN COMMERCE

WASHINGTON, D.C. 20515

June 6, 1980

The Honorable Barley 0. StaggersChairmanCommittee on Interstate and Foreign Commerce2125 Rayburn House Office BuildingWashington, D. C. 20515

Dear Mr. Chairman:

The enclosed report "An Evaluation of the Strategic PetroleumReserve" was prepared by the Congressional Budget Office for theSubcommittee on Energy and Power to evaluate the potential effectivenessof the SPR in time of crude oil import shortfall.

The report is quite specific in its conclusions. When a shortfalloccurs, each barrel of crude oil in the Strategic Petroleum Reserve hasthe potential to prevent almost $200 in current dollars of GNP loss,along with the associated unemployment. Other Congressional sources nowreport that each point of unemployment above the expected base of 6.8percent costs as much as $29 billion in lost revenues and higherunemployment compensation, food stamps and other costs. The cost offilling the reserve is estimated to be approximately $3 billion per year.The costs can be minimized in the long run by expediting the fill as soonas possible. Additionally, the Congressional Budget Office and otherexperts who testified before the Subcommittee indicate there is everylikelihood that, during the period of dependence on imported oil, thiscountry will again experience a substantial shortfall.

Currently, the Department of Energy is not purchasing oil for theStrategic Petroleum Reserve and has not done so since the start of theIranian crisis. Early in 1979, the Administration announced that all newoil procurement would be postponed until after the June 1979 OPEC meetingto avoid price pressure on the world crude oil market. One year later,as OPEC is about to open its 1980 meeting, the same reason is beingrepeated by the Administration as an excuse for not filling the reserve.Additionally, there is speculation that budgetary reasons may also beinfluencing the decision not to fill the Reserve. This report indicatesthat the economic consequences of not filling the reserve are extremelysevere.

(HI)

IV

The Subcommittee appreciates the excellent work performed by theCongressional Budget Office in prejia*4*tt this outstanding report. Widedistribution will assist in maJetfig the logical decision concerning thefuture of the Strategic Petroleum Reserve, and I believe that thisdocument merits publication as a Commjfttee Print.

/Sincerely,

John D. DingellChairman

Enclosure

PREFACE

The Strategic Petroleum Reserve program has experienced numerousproblems in both constructing storage capacity and purchasing oil to fill theexisting capacity. These problems, as well as the current pressures to reducethe federal budget, have raised a number of questions regarding the appropri-ate size of the reserve and the timing of crude oil purchases. This paper,prepared by the Congressional Budget Office (CBO) at the request of theEnergy and Power Subcommittee of the House Interstate and ForeignCommerce Committee, addresses these two major issues.

The report was prepared by Barry J. Holt of the Natural Resources andCommerce Division and Mark Berkman of the Budget Analysis Division, underthe supervision of Raymond C. Scheppach and Robert A. Sunshine. EverettM. Ehrlich and Mark Steitz provided many valuable suggestions. Bill Finan ofWharton Econometrics Forecasting Associates provided technical support.Patricia H. Johnston edited the manuscript. The several drafts were typedby Dorleen Dove, Brice McDaniel, Kathryn Quattrone, and Deborah Vogt. Inkeeping with CBO's mandate to provide objective analysis, the reportcontains no recommendations.

Alice M. RivlinDirector

June 1980

(V)

SUMMARY

The Strategic Petroleum Reserve program, which is administered bythe Department of Energy, has experienced serious difficulties both indeveloping storage capacity and in acquiring oil. Although the Congressauthorized a reserve of 1 billion barrels, only 248 million barrels of storagecapacity have been completed, and only 92 million barrels of oil have beenacquired, the last in August 1979. Furthermore, the President's revisedbudget for fiscal year 1981 reflects a decision not to resume oil purchasesuntil June of that year.

The Congress is now considering legislation that would direct theDepartment of Energy to resume crude oil purchases. On the other hand,concern for a balanced budget in fiscal year 1981 and the possible effects ofrenewed purchases on the international oil market have focused Congres-sional attention on reevaluating the program to determine its size, rates ofoil acquisition, and financing methods.

BENEFITS AND COSTS OF THE RESERVE

The Strategic Petroleum Reserve (SPR) is intended to mitigate theeconomic problems that would be caused by full or partial interruption ofthe flow of oil from abroad. During a supply interruption, the availability ofthe reserve oil would tend to reduce the upward pressure on oil prices andreduce potential losses in gross national product (GNP). For example,without a strategic reserve, a year-long shortfall of 2 million barrels perday in 1984—that is, about 21 percent of projected oil imports—wouldreduce GNP in that year by approximately $146 billion (3.6 percent ofprojected current dollar GNP); it would also increase the unemployment rateby about 1.1 percentage points and the inflation rate by 7 percentage points.This assumes that there would be no price controls or oil allocationregulations. Virtually the entire impact of such a shortfall on the GNP andunemployment, as well as a large portion of the inflationary effect, could beaverted by a reserve of about 750 million barrels. Similarly, a reserve of1 billion barrels could almost completely offset a year-long shortfall of 3million barrels per day in 1984—about 31 percent of projected oil imports—which would otherwise cause about a $226 billion (5.5 percent) loss incurrent dollar GNP and increase the unemployment and inflation rates byabout 1.8 percentage points and 15 percentage points, respectively. The1 billion barrel reserve could also offset the economic effects of a cutoff ofall imported oil for almost four months.

(vn)

VIII

The costs of developing the SPR are not insignificant. Developmentof each Z50 million barrel increment of storage over the current level wouldcost between $1 billion and $5 billion, depending upon the timing, the rate ofconstruction, and the type of storage capacity. Total additional storagedevelopment costs to reach a 1 billion barrel reserve are estimated to rangebetween $5 billion and $11 billion in current dollars over the next six to tenyears. Filling the capacity with the purchase of 1 billion barrels of oil couldbe expected to cost approximately $50 billion, but these costs could berecovered, and probably with a profit, when the reserve was drawn down orthe program terminated.

Each barrel of reserve oil, for which storage capacity costs between$3 and $20 to develop, would thus appear to have the potential to offsetclose to $200 in current dollar GNP losses. Further, although the timing,size, and duration of future interruptions are difficult to predict, the risk ofinterruption seems to be increasing. The level of imports is expected to risefrom the 1979 level of 42 percent of total domestic petroleum consumptionto about 57 percent in 1990, primarily because domestic production isdeclining and demand is not. In addition, the political stability of severalkey producing countries is increasingly uncertain, so that one or more oilsupply interruptions in the next 20 years appear probable. The low cost ofthe oil reserve relative to the economic losses it could avert make thereserve a highly cost-effective federal program for protecting against therisks of growing dependence on imported oil. Indeed, it is the only programthat could offset the short-term economic effects of oil supplyinterruptions.

OIL ACQUISITION STRATEGY

The program's large benefits relative to costs, as well as increasingU.S. dependence on imported oil, seem to bolster the arguments for acquir-ing the oil as rapidly as possible, subject to the constraints imposed by thefederal budget and the international oil market.

Budget Constraints and Offsetting Options

Costs. At projected world oil prices, it would cost approximately$3.0 billion per year over the next few years to purchase oil at a rate of200,000 barrels per day. At this rate, it would take approximately two yearsto fill the current storage capacity of 248 million barrels. If oil werepurchased at a rate of 400,000 barrels per day, this capacity would be filledin 13 months, at a total cost over that period of approximately $5.6 billion.While this represents a sizable outlay, the money spent on purchasing oil isexpected to be recouped when the oil is eventually sold. Moreover, becauseCBO expects continued real increases in the price of oil, the total budgetary

DC

cost of the program might be reduced by expanding capacity and purchasingmore oil sooner, despite the premiums that would have to be paid foraccelerating storage development.

Options to Reduce Budget Effects. Methods could be found toreduce the impact on the budget of spending for the SPR. One would be toimpose an energy-related revenue measure such as the recently proposed oilimport fee and dedicate the revenues to building the reserve. Such a feecould fund the reserve program while, at the same time, discourageconsumption and reduce the level of imports. The Congress could loseflexibility in allocating budgetary resources, however, if this were accom-plished through a trust fund.

Another option would allow private investors to buy title to oil in thereserve, and buy and sell the titles speculatively. In the event of adrawdown of the reserves, the government, retaining control of oil, wouldpay titleholders for the quantity sold.

An additional alternative would be for the federal governmentto tax the subsidy received by refiners of Alaskan North Slope Oil. Thissubsidy results from a combination of the controlled price of the oil and itstreatment under the entitlement program. While oil price decontrol willphase out this subsidy by October 1981, taxing it in the interim could resultin revenues of about $1.6 billion by that time.

International Constraints

International considerations have played a major role in slowing theprogress of the reserve program. Producing nations have opposed theprogram publicly, while agreements among consuming nations apparentlylimit the U.S. government's ability to purchase oil for the reserve in a tight

market.

Producer Opposition. While the public opposition of the producernations to the reserve program is clear, their response to renewed U.S.stockpiling is difficult to predict. The possibility of production cutbackscannot be ignored, especially while countries such as Saudi Arabia areproducing at higher levels than they would prefer. Producer opposition torenewed stockpiling might be minimized, however, if the United Statesdedicated some of the production from federally owned oil to the reserve.One option would be to dedicate the current 130,000 barrels per day federalshare from the Naval Petroleum Reserve (NPR) or some of the royalty oilthat is due the federal government for offshore leases. Although suchdedication would force the refineries currently receiving this oil to replace

63-724 0 - 8 0 - 2

it with purchases on the world market, the strategy might be moreacceptable to producing nations.

Market Restraints. The countries belonging to the InternationalEnergy Agency (IEA) agreed in 1979 to consult with each other beforeresuming stockpiling. They further agreed that no country would resumestockpiling if that would result in pressure on world oil prices. Substantialoil purchases for the reserve in a tight oil market could result in increasedprices that would be felt by all consumers. A high purchase rate of 400,000barrels per day represents about 1 percent of the oil traded daily on theinternational oil market. In a tight, competitive oil market, this could raiseprices as much as $4 per barrel. At lower, more likely purchase rates of100,000 to 200,000 barrels per day, the effect would be only $1 to $2 perbarrel. The oil market, however, does not always exhibit such sensitivity todemand. Prices are not set solely by market conditions. If any priceincreases resulted from SPR purchases, they would more likely stem frompolitical rather than economic factors, and thus would be difficult topredict.

While critics of the reserve have cited tight oil market conditions asa reason for not purchasing oil, the oil market over the next 6 to 18 monthsis expected to be very favorable to reserve purchases. Softening spotmarket prices over recent months, together with the prospects of acontinued recession, may be expected to curtail the rate at which crude oilprices rise over the next year or two.

If market concerns remain a constraint, the United States might wishto consider alternative policies that would have a neutral effect on theworld oil market. For example, it might be possible to increase NPRproduction at Elk Hills by 30,000 to 50,000 barrels per day and dedicate thisadditional production to the reserve. Another alternative might be to findways of reducing current domestic demand and use the resulting oil savingsfor the reserve.

CONCLUSION

The Strategic Petroleum Reserve is a highly cost-effective programfor reducing the risks of high oil import levels and future interruptions ofsupply. To fill the remaining storage capacity of 156 million barrels wouldcost approximately $5.6 billion if it were accomplished in 13 months at ahigh acquisition rate of 400,000 barrels per day. Budget limitations andinternational oil market conditions are the two major constraints to currentpurchases of oil for the reserve. If budget limitations are severe, variousalternatives are available—such as imposing an oil import fee or raising

XI

private financing. With respect to the international oil market, theconstraints are twofold: the opposition of producing nations and possibleupward pressure on oil prices resulting from the reserve purchases. Pro-ducer opposition might be minimized by dedicating federal oil from theNaval Petroleum Reserve or royalty oil from offshore leases. It is possiblethat the oil market will have a considerable amount of slack over the next 6to 18 months, so that the economic climate for renewed purchases should befavorable. If market constraints are serious, however, alternative policiescould be considered, such as increasing oil production from Elk Hills orreducing domestic demand and using the savings for the reserve.

CONTENTS

Page

PREFACE V

SUMMARY VII

CHAPTER I. INTRODUCTION 1

CHAPTER H. EVALUATION OF ALTERNATIVERESERVE SIZES 3

BENEFITS OF A STRATEGIC RESERVE . . . 3

Methodology for Calculation of Benefits . . . 4Estimated Benefits 4

STRATEGIC RESERVE COSTS 7

EVALUATION OF COSTS AND BENEFITS . . 7

CHAPTER m. ALTERNATIVE PURCHASE STRATEGIES . . 11

BUDGET CONSTRAINTS 11

Budget Effects 11Policy Options to Reduce

Budget Effects 13

WORLD OIL MARKET CONDITIONS . . . . 14

Relations with Other Nations 14Policy Options to Minimize Political

and Supply Constraints 15

CONCLUSION 17

APPENDIX A. DESCRIPTION OF THE WHARTONANNUAL ENERGY MODEL 21

APPENDIX B. FOUR RESERVE COMPLETION CASES . . . 31

XIII

TABLES

XIV

Page

TABLE 1. MACROECONOMIC INDICATORS DURINGA FULL YEAR OF OIL SUPPLY INTER-RUPTION IN 1984 6

TABLE 2. DEVELOPMENT COSTS AND EXPECTEDBENEFITS OF A STRATEGICPETROLEUM RESERVE 9

APPENDIX TABLES

Page

TABLE A-l. SIMPLE INPUT-OUTPUT ACCOUNTINGEXAMPLE 22

TABLE A-2. SIMPLE INPUT-OUTPUT ACCOUNTING EXAMPLEWITH EXPANDED FINAL DEMAND 23

TABLE A-3. SIMPLE INPUT-OUTPUT TABLE INCOEFFICIENT FORM 24

TABLE B-l. CASE ONE—DOE SCHEDULE 32

TABLE B-2. CASE TWO—ACCELERATED OILACQUISITION 32

TABLE B-3. CASE THREE—ACCELERATED STORAGEDEVELOPMENT AND OIL ACQUISITION . . . 33

TABLE B-4. CASE FOUR—RELIANCE ON FEDERALLYOWNED OIL 34

TABLE B-5. SUMMARY OF COSTS FOR A 728MILLION BARREL RESERVE 35

XV

FIGURES

Page

FIGURE A-l. BASIC INPUT-OUTPUT STRUCTURE . . . . 25

FIGURE A-2. DIRECTION OF SOLUTION OF INPUT-OUTPUT RELATIONSHIPS 25

FIGURE A-3. DIRECTION OF SOLUTION OF PRICERELATIONSHIPS 26

AN EVALUATION OF THE STRATEGIC PETROLEUM RESERVE

CHAPTER I. INTRODUCTION

The security of the U.S. oil supply has been a major goal of energypolicy since the 1973-1974 embargo by the Organization of PetroleumExporting Countries (OPEC). To protect the country from the consequencesof future interruptions in the supply of imported oil, the Congress autho-rized a federal petroleum reserve in the Energy Policy and Conservation Actof 1975 (EPCA). In June 1978, the Congress approved plans for a StrategicPetroleum Reserve (SPR) of 1 billion barrels, the maximum amount autho-rized by the EPCA. The potential usefulness of this program has grownalong with the rising level of imported oil—now about equal to domesticproduction—and increasing political instability in the oil-producing nations.

The SPR program, administered by the Department of Energy (DOE),has experienced serious difficulties both in developing storage capacity andin acquiring oil to meet the authorized level of 1 billion barrels. To date,storage capacity of only 248 million barrels has been completed, withanother 290 million barrels of capacity now under construction. Develop-ment of the remaining 462 million barrels of capacity has been delayed until1982, both because of oil supply uncertainty and because using privatesector storage has proved to be very expensive. Furthermore, while most ofthe technical problems in preparing storage facilities have been solved, thereserve today contains only 92 million barrels. \j The reserve program hasnot contracted for more oil since February 1979 and has taken no deliveriessince August 1979.

Thus, the completion of a 1 billion barrel reserve is by no meanscertain. The Energy Security Act of 1979 (now in conference) would directDOE to resume the purchase of oil. On the other hand, concern for abalanced budget in fiscal year 1981 has also focused Congressional attentionon reducing or eliminating the program. Of the $6.9 billion that wasappropriated for the program between fiscal years 1976 and 1980, about

\J To date all storage facilities are salt domes and mines. The programencountered a number of technical difficulties, including limited ratesof leeching for brine removal, unanticipated environmental restrictionson brine disposal, and equipment failure.

(1)

63-724 0 - 8 0 - 3

$4.1 billion will still be available at the beginning of fiscal year 1981. 2JThe President's January budget request for fiscal year 1981 devoted16 percent of the outlays for energy (function 270) to the SPR, and calledfor the resumption of oil acquisition in June, 1980, at the rate of 3 millionbarrels per month, increasing to 7.5 million barrels per month in calendaryear 1982. The revised budget request submitted in March, however, delayspurchases until June 1981. Recent statements by DOE officials indicatethat no final decision regarding oil purchase has been made at this time.

To resolve these issues, the Congress is again evaluating the reserveprogram to determine its size, rates of oil acquisition, and financingmethods. Chapter n assesses the value of a strategic reserve in mitigatingthe effects of an oil supply interruption. The reserve's value is comparedwith its costs to evaluate various reserve sizes. Chapter in identifies thosefactors that constrain the rate at which storage capacity can be built and oilacquired. Policy options are suggested for reducing these constraints andallowing high rates of acquisition. Two appendixes provide additionalinformation. The first is a detailed description of the model used tocalculate the benefits of the reserve. The second compares the budgeteffects of alternative storage construction and oil acquisition strategies.

2/ Approximately $1.8 billion has been committed to the Department ofDefense's Defense Fuel Supply Center, but DOE has not authorized theCenter to purchase oil. The remaining $2.3 billion is presently unobli-gated and requires reappropriation by December 31, 1980.

CHAPTER H. EVALUATION OF ALTERNATIVE RESERVE SIZES

The United States imported about 42 percent (8.0 million barrels perday) of its total oil supplies in 1979. With relatively stable consumptionlevels and diminishing domestic production levels, this proportion is likely toincrease to 57 percent by 1990.

The U.S. dependence on oil imports, which are vulnerable to supplyinterruptions, poses a distinct set of risks for the U.S. economy. These risksmay be identified as economic losses or costs that are not included in theprice of oil. They are of uncertain magnitude and related to events thatmay, but not necessarily will, occur. This reliance on uncertain importsconstrains foreign relations and leaves the economy vulnerable to poten-tially significant disruptions. !_/

There are two basic ways to limit the possible effects of an oil supplyinterruption: reducing the dependence on foreign oil and preparing tomitigate the short term effects of an interruption. Imports can be reducedby substitution of alternative energy sources, by new domestic production orby conservation. Such policies, however, affect long-term production andconsumption patterns and offer very little immediate protection. In theshort term, the effects of a supply interruption can be alleviated bysufficient domestic petroleum reserves as well as by effective rationing andallocation programs. This chapter evaluates various oil reserve sizesdesigned to provide this short-term protection. Reserves ranging between250 million barrels and 1 billion barrels are compared in terms of estimatedbenefits and costs.

BENEFITS OF A STRATEGIC RESERVE

The most significant measurable benefits of a strategic reserve arethe adverse macroeconomic effects that it would mitigate in the event of anoil supply interruption. Drawing oil from a strategic reserve during supplyinterruptions would allow continued production of many goods and servicesthat would otherwise be curtailed, and would reduce surges in prices andunemployment. The value of the strategic reserve—that is, the expected

\J For additional information, see Congressional Budget Office, The WorldOil Market in the 1980s; Implications for the United States, Back-ground Paper (May 1980).

(3)

benefits to be derived from the reserve—depends on the probability, size,duration, and timing of future supply interruptions. While there is muchuncertainty in any estimates of such probability, the value of a reserve canbe estimated under a number of scenarios. In the analysis presented here,the macroeconomic effects of an oil supply interruption are explored. Thedegree to which a reserve could offset these effects determines the value ofthe reserve.

Methodology for Calculation of Benefits

The economic impact of an oil supply interruption was calculated interms of its effect on aggregate national output (as measured by the grossnational product), inflation, and unemployment. Estimates were developedusing the Wharton Annual Energy Model, which combines input-outputanalysis with macroeconomic analysis.

This model utilizes input-output tables that relate oil consumption andindustrial output to calculate the effects of oil supply interruptions. Bycomparing the constrained feasible outputs with those required by the newdemand composition, allocation adjustments are determined. This procedureis repeated until the final demands are consistent with the constrainedoutput, resulting in new values for expected gross national product (GNP)and the level of prices and employment. The model allows inflationarysurges from excess demand to be passed through to the general price levelor to be suppressed by modelling the policy device of price controls. Themodel and the assumptions concerning GNP growth, oil prices, demand, andconservation are discussed in more detail in Appendix A.

Estimated Benefits

Scenarios were designed with varying levels of oil supply interruptionsand policy responses, and used in the model. In all cases, the interruptionwas assumed to occur in 1984 and to last for one year. The year 1984 wasselected because it appeared to be the first year when a reserve ofsignificant size (between 250 and 700 million barrels) could reasonably beexpected to exist. This analysis indicates that each barrel of oil provided bythe reserve would avert about $200 of the 1984 GNP loss resulting from aninterruption. In addition, the reserve would suppress surges in the rates ofinflation and unemployment.

Table 1 .summarizes the results of the simulations. 2/ In all cases,real output would decline and unemployment and inflation rates increase asa result of oil supply interruptions. 3/ Higher prices would reduce both realincome and real wealth of households, further reducing purchases andslowing real economic growth. In the years following the oil supplyinterruption, the economy would rebound but would not recover entirely tothe level it would have reached without an interruption. The economiceffects of an oil supply interruption would be severe. In the range ofshortfalls considered here, from 1 to 5 million barrels per day, GNP losswould be slightly greater than proportional to the shortfall. For example, ayear-long, 2 million barrel per day shortfall would result in a GNP loss ofabout $146 billion (3.6 percent of 1984 GNP), and a 3 million barrel per dayshortfall, $226 billion (5.5 percent).

In this analysis, a petroleum reserve would, simply reduce the dailyshortfall and thus reduce the economic effects of the .interruption. 4/ Forexample, in the absence of a reserve, the year-long supply interruption of3 million barrels per day would result in the GNP loss of $226 billion, andincrease Inflation and unemployment by 15 and 1.8 percentage points,respectively. A 750 million barrel reserve would effectively reduce theshortfall to about 1 million barrels per day, and reduce the impactsaccordingly. In this case the reserve would avert $160 billion of the GNP

2/ The macroeconomic impact estimates of an interruption presented inthis paper assume no price controls and an allocation of petroleum thatminimizes GNP losses. This method of allocation differs from thatprovided for by the Energy Policy and Conservation Act (EPCA) of1975. The EPCA rules allocate supplies in a way which would minimizeGNP losses only after the demands of public safety, health, farms,transportation, and utilities are met. The economic losses associatedwith the EPCA allocation rules are high, resulting in a value of thereserve oil considerably higher than those presented in the text. Forexample, a year-long, 3 million barrel per day shortfall would result ina 1984 GNP loss of about $400 billion, as compared to the $226 billionloss under the efficient allocation assumed for the analysis.

3/ The analysis does not specifically account for the costs to the federalgovernment associated with the high rates of inflation. and unemploy-ment that would prevail during an oil supply interruptions.

4/ The simplifying assumption that results in total drawdown of thereserve does not recognize the possiblity of a more cautious drawdownplan reflecting the uncertain duration of supply interruptions. Analysisof alternative drawdown strategies is beyond the scope of this paper.

TABLE 1. MACROECONOMIC INDICATORS DURING A FULL YEAR OF OIL SUPPLYINTERRUPTION IN 1984

DailyShortfall(millionbarrels

per day)

12345

GNP as a LossPercent ofProjectedImports

10.521.131.642.152.6

In Billionsof Dollars

66146226306387

Percent ofProjected

GNP

1.63.65.57.59.4

Increase inProjectedInflation

Rate(percentage

points)

37

152531

Increase in Minimum ReserveProjected Required to

Unemploy-ment Rate(percentage

points)

0.51.11.82.22.8

Offset EntireInterruption(millions of

barrels)

365730

1,0951,4601,825

loss, and reduce inflation and unemployment by 12 and 1.3 percentagepoints, respectively.

It is possible that not all of these benefits would be realized. Foreigncountries might capture many of the benefits of a drawdown of theStrategic Petroleum Reserve. While coordinated stockpiling and drawdownamong consumer nations would lead to large benefits for all consumingnations, it is possible that certain countries could benefit directly from thereserves of another.

The IEA Agreements that allow for the allocation of supplies duringinterruptions would not be triggered unless there were a 7 percent disruptionin the supply for member countries. 5/ In a smaller disruption, withoutcoordinated drawdown, oil drawn from any member's stockpile would beconsidered an increase in world supply and, in principle, would be distributedaccordingly. Although the oil from the reserve would physically remain inthe country, it would to some degree displace oil that would have beenimported. If oil from the Strategic Petroleum Reserve were distributedthroughout the world market in this way, as much as two-thirds could "leak"to other countries. Consequently, if there were no coordination of reservedrawdown among consumer nations, larger SPR size and faster drawdawnrates would be needed to achieve the same results.

5/ Such a disruption would leave the United States facing a shortfall of 1.0to 1.5 million barrels per day.

STRATEGIC RESERVE COSTS

Although the budgetary impact of the reserve program includes boththe costs of storage facility development and oil acquisition, only develop-ment costs are compared here to the benefits identified in the previoussection. While-oil acquisition rcosts account for over 80 percent of thebudgetary costs, the oil should maintain its value or, in all likelihood,increase before H is^used. Consequently, in this analysis, the purchase of oilis not considered a cost, because the expenditures will be recovered uponreserve drawdown.

The costs to complete 1 billion barrels of capacity range from$5 billion to $11 billion over the next ten years, depending on the type ofstorage capacity used and when the capacity is developed. The program nowrelies upon salt domes and, in one case, a salt mine to store oil. All of thefirst 538 million barrels will be stored in this fashion. Additional storagecould be developed through a variety of methods, including additional saltdomes and mines as well as above-ground steel tanks.

The Department of Energy currently estimates that the first 538million barrels of capacity will cost approximately $3 per barrel or $1.6billion. The Department estimates that the next addition of 190 millionbarrels of capacity will cost approximately $5 per barrel or $1 billion. NoDOE estimate is available for the costs of completing the last 272 millionbarrels of capacity. Assuming costs for this increment are similar to thoseof earlier increments, storage costs would total $5 billion for the 1 billionbarrel reserve. Reliance on private sector storage, however, could result insignificantly higher costs. Initial bids from the private sector ranged from$4 to $15 per barrel. 6/ Total development costs thus could be as high as$11 billion if the upper bound of this range is assumed for the 462 millionbarrels of capacity not already under construction.

Withdrawal costs are estimated to be between $20 million and $100million. This cost will depend on the level of withdrawal and could be borneby the purchasers of the oil.

EVALUATION OF COSTS AND BENEFITS

A direct comparison of development costs to the benefits of astrategic reserve must be made carefully. The costs represent federal

6/ These were for 20-year leases of salt domes, mines, and steel tanks.

outlays over a short period of time for a program that could provide securityfor many years. The quantifiable benefits of the reserve are measured interms of offsetting economic losses in the event of future supply interrup-tions. As shown in Table 2, expected economic loss offsets are exceedinglylarge. For example, if the country were to face a single year-long supplyinterruption of 3 million barrels per day, a reserve of 1 billion barrels wouldavert about $205 billion of GNP loss, compared to development costs ofbetween $5 and $11 billion. 7/

A cost-benefit evaluation of the reserve is difficult without knowingthe likelihood of future oil supply interruptions. The probability of futuresupply interruptions is, however, difficult to estimate. The oil distributionsystem can be disrupted by logistical problems, political instability, ormilitary action. Historical evidence suggests that the probability ofinterruptions is not low. Since 1973, when OPEC first began to restrictoutput, there have been two supply interruptions, the political embargo of1973-1974 and the interruption resulting from the Iranian revolution begin-ning late in 1978. Since the UN resolution establishing Israel in 1947, therehave been at least four other interruptions of various sizes, including twoblockades of the Suez Canal, a bombing of a pipeline in Iraq, and a fire at anoil field in Saudi Arabia. While these interruptions have not all had seriousworldwide consequences, they have occurred with an average frequency ofonce every five to six years since 1947.

There are many situations today that could result in the UnitedStates facing shortfalls of up to 5 million barrels per day. Politicalinstability and the possibility of military activity continue to pose a threatto oil market stability. Strained relations between Iran and Iraq, or theUnited States and Iran, could escalate and result in severe market disrup-tions. An abrupt, uncompensated export cutback by either Nigeria or Libya,for example, could result in a shortfall to the United States of about1 million barrels per day. A general OPEC cutback of about 20 percentwould cause a 2 million barrel per day shortfall to the United States.Logistical problems or military action at one of the centralized points in theworld oil delivery system could result in much larger disruptions. Forexample, if the Strait of Hormuz were closed, the United States could face

7/ Because the program benefits are measured in dollars of GNP and theprogram costs are measured in dollars of federal expenditure, they maynot be directly comparable on a net basis. This is because one dollar offederal expenditure may result in an increase of more than one dollar inGNP. This makes the value of net benefits somewhat uncertain. Giventhe magnitude of the difference between costs and benefits, however,this does not appear to significantly change the results.

TABLE 2. DEVELOPMENT COSTS AND EXPECTED BENEFITS OF ASTRATEGIC PETROLEUM RESERVE

SPR Size Development Costs a/ Benefits a/(millions of barrels) (billions of dollars) (billions of dollars)

250 1 b/ 45500 2-3 95750 3-7 150

1,000 5-11 205

a/ Benefits are measured in terms of averted GNP loss in the event of ashortfall that would require drawdown of the entire reserve in 1984 andare in 1984 dollars. The costs are in current dollars, incurred bothbefore and after 1984. Adjusted to 1984 dollars, the costs of smallerreserve sizes would increase slightly while costs for the larger reservewould decrease. This does not significantly change the results.

b/ At the end of fiscal year 1980, approximately $0.9 billion will have beenspent for the first 248 million barrels of storage capacity.

a shortfall of 5 million barrels per day or more. These situations andhistorical evidence, then, suggest that the United States is likely to face oneor more supply interruptions during the next two decades.

Table 2 shows that the probability of an interruption need not belarge to make expected benefits of the 1 billion barrel reserve greater thanprojected costs. The net benefits expected from the reserve increase withthe size of the reserve with any fixed probability of total drawdown.Further, using the high end of the cost range, the ratio of benefits to costsof the 1 billion barrel reserve indicates that it would be economic if theprobabilities of an interruption were such that complete drawdown isexpected during the next 19 years.

CHAPTER m. ALTERNATIVE PURCHASE STRATEGIES

Chapter IE indicates that the reserve program, with the Congres-sionally approved 1 billion barrel goal, would be cost effective. The soonerthe reserve is operational, the more protection it will provide. Budgetconstraints and international oil market conditions, however, restrain therate at which storage capacity can be built and oil acquired. These factorsare discussed here, along with policy options for addressing them.

BUDGET CONSTRAINTS

Budget Effects

The Congress has appropriated $6.9 billion for the SPR to date. Atthe end of fiscal year 1979, approximately $Z.l billion had been spent,leaving $4.8 billion unexpended ($2.3 billion will require reappropriation bythe end of calendar year 1980). An additional appropriation of at least $1.5billion would be required to complete a 250 million barrel reserve and $55billion or more would be necessary to complete a 1 billion barrel reserve.Outlays for the reserve would also be a significant portion of annual federalenergy expenditures. In the President's January budget for fiscal year 1981,SPR outlays accounted for 16 percent of budget outlays for energy. TheSPR outlays would grow to 35 percent by 1984 under the Department ofEnergy's January budget plan. I/ This percentage could be even highershould a more aggressive development and oil acquisition schedule beundertaken.

Timing is a critical factor in determining the ultimate budget impactof the reserve program. Generally, the federal government could minimizebudget costs by buying oil as soon as possible. CBO currently assumes oilprices will increase in real terms by 2 to 3 percent annually over the nextten years. Savings of this kind, however, might be at least partially offset ifa premium would have to be paid in order to accelerate the development ofstorage capacity. Four development and oil acquisition scenarios weredeveloped to demonstrate the budget impact of alternative completionschedules and the trade off between oil acquisition savings and developmentcost increases. The scenarios, discussed in detail in Appendix B, include the

V The President's March update to the fiscal year 1981 budget calls for aone-year delay in oil purchases, from June 1980 to June 1981.

(11)

12

DOE schedule as presented in the January Budget request (Case One); aschedule filling existing capacity as quickly as possible (Case Two); aschedule accelerating storage capacity development as well as oil acquisi-tion (Case Three); and a schedule linking oil fill to production levels offederally owned oil reserves (Case Four).

This comparison indicates that the use of more expensive facilities toaccelerate the program does not necessarily increase total program costs.The premium that must be paid for more rapid development of facilitiesmay be more than offset by savings realized from the earlier purchase ofoil.

The Department of Energy schedule, defined in its fiscal year 1981budget request submitted in January, calls for resumption of oil acquisitionin June 1980 at a rate of 100,000 barrels per day. The rate of acquisition isto remain at this level until the beginning of calendar year 1982, when it isincreased to 250,000 barrels per day. This is not the maximum rate of fillthat could be accomplished. The maximum rate of fill at existing SPR sitesis approximately 400,000 barrels per day. At this rate, existing remainingcapacity of 156 million barrels would be filled within 13 months after theresumption of oil acquisition.

Construction of additional storage of 290 million barrels is scheduledto continue through fiscal year 1981 by expanding existing sites. Nodecision regarding the next expansion has been made. The January budgetrequest, however, includes planning funds for further expansion, includingregional storage of 24 million barrels, primarily for the Northeast andHawaii. Storage capacity of 728 million barrels is projected by 1990. Itmay be possible, though, to accelerate development in order to takeadvantage of oil market conditions and to maximize reserve benefits, inspite of the higher development costs.

To expand capacity more rapidly, the government would have tointensify federal efforts or increase the use of the private sector. One ofthe approaches DOE has considered, the "turnkey" approach, calls for thelease of storage capacity, above or below ground, from private firms. TheDepartment has not proceeded with this approach for two reasons. Firstbecause of uncertainty of oil availability, the long-term leases requiredwere not thought advisable. Second, initial contract bids were higher thanDOE anticipated and significantly higher than estimated costs of govern-ment facility expansion. A willingness to pay a premium for accelerateddevelopment would allow greater private sector involvement or more costlyabove ground storage. Such a policy could result in additional capacityleading to a total of 450 million barrels in 1982 and 1 billion in 1986. The

13

January DOE plan calls for only 300 million barrels of capacity in 1982 and560 million by 1987. 2/

Policy Options to Reduce Budget Effects

Options to reduce the net budget impacts of the Strategic PetroleumReserve include:

o Dedication of a portion of the revenues from a new, energy-related revenue source to the reserve.

o Private financing through a mechanism that allows the specula-tive trading of titles to reserve oil.

o Reallocation of the subsidy provided by the entitlement treatmentof Alaskan North Slope oil.

New Revenue Sources. There are a number of options that thefederal government could employ to raise revenues for the reserve programand minimize the net budget impact. The recently proposed oil import feeand the excise tax that was to replace it represent efforts to discourageconsumption and to reduce the long-term risks associated with high levels ofimports. It might be possible to impose such a fee for the purpose offunding the reserve program. Any formal creation of such a trust fund,however, could reduce Congressional flexibility in allocating budgetaryresources.

Private Financing of the Strategic Reserve. If budgetary constraintsinhibit the filling of the Strategic Reserve, private financing might besought. This could be accomplished by allowing investors to purchase atransferable title to reserve oil. The federal government would retaincontrol of the oil but allow the speculative trading of the titles.

In the event of a drawdown of the reserve, the federal governmentcould prorate receipts among all titleholders. In order to prevent pricemanipulations at the time of sale, the government could agree to charge theprice quoted for a reserve barrel title on. the day that the President decidedto draw down the reserve. A consequent refilling of the reserve could beaccomplished by issuing a new set of titles on the open market.

2/ No change in the development schedule was made in the March 1980revision to the fiscal year 1981 budget request.

14

Alaskan North Slope Entitlements. Another source of funds might befound in the treatment of Alaskan North Slope (ANS) oil in the entitlementsprogram. 3/ Refiners of Alaskan crude have been provided a subsidy, at theexpense of refiners and consumers of other domestic oil, because of ANSoil's unique regulatory position prior to the OPEC price increases of 1979.At that time, refiners were granted entitlements for ANS oil because thesum of wellhead price and transportation costs were higher than the worldprice for oil. The controlled wellhead price of North Slope oil, plustransportation costs, is now below the world price. The refiners of ANS oil,however, still do not have to buy the right to refine the oil. This implicitsubsidy is currently valued at about $5 per barrel, which will be reducedthrough 1980 and 1981 as the wellhead price is decontrolled. If the federalgovernment were to tax away the subsidy, approximately $1.6 billion couldbe realized between now and October 1981. This money could then bededicated to the reserve. DOE is currently exploring alternatives to removethe advantage that refiners of ANS oil receive from its entitlements-freetreatment.

WORLD OIL MARKET CONDITIONS

International considerations have played a major role in slowing theprogress of the reserve program. Agreements with other members of theInternational Energy Agency, based on tight market conditions, apparentlylimit the government purchases of oil for the reserve. In addition, theproducing nations have opposed the program publicly.

Relations with Other Nations

Consuming Nations. The IEA .member countries agreed in 1979 toconsult with each other prior to continued stockpiling. They further agreed

3/ The entitlements program equalizes the cost of crude oil for alldomestic refiners, through the buying and selling of entitlements(defined as the right to refine a barrel of controlled domestic oil).When a refiner refines a barrel of imported oil, he is allowed to sell anentitlement to an entitlements pool. When refining a barrel ofcontrolled domestic oil, he is compelled to buy an entitlement. Thevalue of these entitlements is continually adjusted so that the buyingand selling leaves the entitlements pool with a final value of zero. Byforcing refiners of less expensive domestic oil to buy the right to refineit, and by allowing refiners of more expensive imported oil to sell theirunused entitlements, the system equalizes the average cost of crude oilfor each refiner.

15

that no country would resume stockpiling efforts, if such efforts wouldresult in pressure on the world oil market.

High rates of oil purchase for the reserve in a tight oil market couldresult in increased prices that would not be limited to SPR oil, but would befelt by all consumers. A high purchase rate of 400,000 barrels per dayrepresents about 1 percent of the oil traded daily on the international oilmarket. If the oil market were tight and truly competitive, such a purchaserate could increase prices as much as $4 per barrel. 4/ The world oil marketdoes not always exhibit such sensitivity, however, and prices are not setsolely by market conditions. Price increases, if any, resulting from SPRpurchases are more likely to be politically motivated and, consequently,impossible to predict.

The world situation may never be completely favorable to the federalstockpiling of oil. Current market conditions and the quantities of oilrequired for the reserve program, however, suggest that resumed stockpilingmight not have adverse effects on the market. Moreover, softening spotmarket prices over recent months and the prospects of a continued recessionthat would suppress demand indicate that the next 6 to 18 months maypresent a relatively good opportunity to acquire substantial quantities of oilwithout significantly increasing oil prices.

Producing Nations. While the position of those producer nationspublicly opposing the reserve program is clear, their response to renewedstockpiling is difficult to predict. The possibility of production cutbackscannot be ignored, especially while countries such as Saudi Arabia areproducing at levels above their stated preferences. At the same time,however, some of the smaller producers that rely on spot markets whichhave been weakening might be willing to negotiate contracts to supply oil tothe reserve. After the OPEC meeting in June 1980, however, certainproducers, including Saudi Arabia, may reduce production levels regardlessof the status of the SPR program.

Policy Options to Minimize Political and Supply Constraints

The above factors notwithstanding, there are policy options thatmight minimize the political and supply constraints on the SPR. The optionsto minimize political constraints include the use of federally owned oil anddiplomatic or trade agreements. Implementing demand reducing policies or

4/ The price of oil is assumed to be $40.00 per barrel. The price elasticityassumed is -O.I.

16

increasing production of federally owned oil could ease the constraints posedby a tight oil market.

Policies to Reduce International Political Constraints. Internationalconstraints, caused by producer opposition to the reserve and consumerfears of tight markets and higher prices, might be reduced through the useof federally owned oil to fill the reserve. Cutting off the sale of this oil andearmarking it for the strategic reserve would force its present customers tothe world market, however, and thus have the same net effect on oilsupplies as open federal purchases. Nevertheless, such policies obviate theneed for the government to enter directly into the international oil market.

The federal government currently sells about 130,000 barrels per dayfrom the Naval Petroleum Reserve (NPR), some of which could be tradedfor oil to be placed in the Strategic Petroleum Reserve. Adverse publicreaction to the recent sales at high market prices as well as efforts bypurchasers to renege on their contracts provide incentives to use NPR oilfor the strategic reserve. There are, however, a number of West Coastrefiners that would have trouble replacing this high quality oil or adapting toheavier grades. Although using NPR oil has been cited as a means ofreducing the budgetary cost of the SPR, such use would not accomplish this.While SPR outlays would be reduced by the oil transfer, the loss of NPRreceipts would offset this outlay reduction.

A second major source of federally owned oil exists in the Gulf ofMexico, in the form of royalty oil due the federal government for offshoreleases. This royalty oil is now sold back to the producers or, in some cases,distributed to other refiners. There is approximately 100,000 barrels perday of new production each year, of which about 15,000 barrels per day isowed the government. Policies could be implemented to accept all newroyalty oil in kind rather than cash, and dedicate it to the strategic reserve.This would result in an incremental acquisition of 15,000 barrels per dayeach year. In addition, natural gas leases in the Gulf produce about 3.6trillion cubic feet per year, yielding the equivalent of about 30,000 barrelsof oil per day in royalty gas. Some portion of this gas could be traded for oilto be placed in the SPR.

International political constraints might also be reduced throughspecial trade agreements with oil-producing nations. Oil in exchange forU.S. technology might prove a highly effective form of trade. 5/ SaudiArabia, for example, might be willing to assure the supply of oil to the

5/ Congressional Budget Office, The World Oil Market in the 1980s;Implications for the United States, Background Paper (May 1980).

17

United States in exchange for stronger military support. Less developedoil-producing nations might prefer other types of trade agreements. Someof these agreements, however, might address the long-term problem of oilsecurity more effectively than solving the short-term needs of the reserveprogram.

Policies to Minimize Oil Supply Constraints. While utilizing federalsources of oil for the reserve might be politically more acceptable toproducing and consuming nations, it would still result in additional demandfor oil. It might be possible, however, to increase production from thesefederally owned sources and dedicate the increment to the reserve. Forexample, the NPR reserve at Elk Hills, California, could produce anadditional 30,000 to 50,000 barrels per day and stay within the estimatedmaximum efficient rate.

Another way to obtain additional oil for the reserve without dis-turbing the world oil market would be to offset the additional demand byreducing consumption for other uses. Thus the reserve could be filledwithout increasing the present level of imports. Current and proposedpolicies, such as utility coal conversion directives and conservation incen-tives, are aimed primarily at reducing the demand for oil over the nextdecade and beyond. To make oil available for the SPR, more immediatedemand restraint would be needed. This restraint could be achieved bystringent policies, such as gasoline rationing or mandated conservation, orby increased economic incentives for rapid fuel switching, increased energyproduction, and conservation.

CONCLUSION

This chapter indicates that there are constraints on the rate at whichthe Strategic Petroleum Reserve can be developed. While these constraintscould be reduced or eliminated, there are tradeoffs that must be considered.Total long-term budget effects could be reduced by accelerated storageconstruction and oil acquisition, though short-term budget effects would beincreased. Short-term budget effects could be addressed through the use ofprivate sector funds, the dedication of new energy-related revenues, orchanging the entitlement treatment of Alaskan oil.

Earlier reserve completion would provide greater program benefitsthrough earlier protection against supply interruptions, although the morerapid oil purchases might create international tension. Current marketconditions and the relatively small levels of oil purchases required by theprogram suggest that these international concerns might not present aserious constraint, however. Finally, changes in domestic and foreignenergy policy might provide additional oil supplies that could be dedicatedto the reserve.

APPENDIXES

APPENDIX A. DESCRIPTION OF THE WHARTON ANNUALENERGY MODEL

Introduction

Input-output analysis offers an effective technique for studying therelationship between the size of an oil supply interruption and aggregateeconomic activity. This type of analysis captures both the direct andindirect effects of an oil shortfall.

In this application, however, conventional input-output analysis hassome disadvantages. First, conventional input-output analysis assumes fixedtechnologies and prices—that is, no input substitution is allowed. Second,while permitting the determination of industrial deliveries to final demand,it does not determine the distribution of these deliveries by GNP com-ponent. Third, input-output analysis does not recognize inherent outputrestrictions that exist in such industries as oil and gas.

These problems can be resolved by combining conventional input-output analysis with a framework that permits input substitution, distributesindustrial deliveries to final demand across GNP components, and restrainsoil and gas output. The energy version of the Wharton Annual Modelcombines input-output analysis with a macroeconomic model, eliminatingmany of the problems inherent in conventional input-output analysis.

To determine the impact of petroleum import shortfalls, the WhartonAnnual Energy Model (WAEM) is "solved backwards." Restrictions onindustrial output restrain the gross output of the model's 63 industrialsectors. The assignment of priorities to each industry in WAEM is based onstudies done for the Federal Energy Administration by Resource PlanningAssociates (RPA) of Cambridge, Massachusetts. The input-output table isthen used to determine the feasible set of deliveries to final demand. Bycomparing the feasible set of deliveries to final demand with the setrequired by the existing final demand composition, necessary allocationadjustments to final demand are determined. The purpose of theseadjustments is to eliminate excess demand for the output of each industrialsector. This procedure is repeated until the final demands are consistentwith the constrained output.

(21)

22

The presence of excess demand in the system is an obvious source ofinflation during an oil production cutback. The model is structured to allowthese inflationary surges either to be passed through to the general pricelevel and wages or, alternatively, to be suppressed. The latter option positsthe existence of price restraints, such as controls, to prevent primary pricesfrom adjusting because of excess demand.

Input-Output Analysis of a Foreign Oil Production Cutback

Consider an input-output model of a fictitious three-industryeconomy consisting of agriculture, steel, and oil. In Table A-l each columnprovides a list of the inputs needed for the production of an industrial good.For example, for each $100 of oil output, $25 of input is required from thesteel industry, and $65 in labor and capital inputs. Each row gives thedistribution of an industry's output. Again, for the oil industry, $25 isdelivered to the agricultural industry, $50 to steel, $10 to oil, and $15 tofinal demand. For each industry, the row and column sums equal the totalgross value of output for that sector.

TABLE A-l. SIMPLE INPUT-OUTPUT ACCOUNTING EXAMPLE (In dollars)

IndustrialDeliveries to Gross

Agriculture

Steel

Oil

Value Added a/

Gross Output

Agriculture

50

25

25

50

150

Steel

0

100

50

50

200

Oil Final Demand Output

0

25

10

65

100

100

50

15

165

—

150

200

100

450

NOTE: GNP = value added = J industrial deliveries to final demand= 165.

a/ Labor and capital inputs.

23

Now, consider the same example, with the exception that the singlefinal demand category is broken down into three categories: consumption,net exports, and other final demands (see Table A-2). In this case, both thesummation of the value added and the summation of the GNP componentsyield GNP of $165.

TABLE A-2. SIMPLE INPUT-OUTPUT ACCOUNTING EXAMPLE WITH EXPANDEDFINAL DEMAND (In dollars)

Industrial DeliveriesIntermediate Inputs to Final Demand

Agriculture

Agriculture

Steel

Oil

Value Added a/

GNP Components

Gross Output

50

25

25

50

150

Steel

0

100

50

50

200

Oil Consumption

0 80

25* 0

10 10

65

— 90

100

Net GrossExports Other Output

10 10 150

50 0 200

5 0 100

—

65 10

450

NOTE: GNP = I value added = I GNP components = 165.

a/ Labor and capital inputs.

If the intermediate sector column entries are divided by the grossoutput of the sector, the proportions for $1 of gross output are obtained (seeTable A-3). Similarly, if each component of final demand is divided by itscolumn total, the industrial distribution of $1 of each category of finalexpenditure is obtained, which yields Table A-3.

Figure A-l shows the basic structural elements of the simple tablesin this appendix. In solving the Wharton Annual Model, a vector of GNPcomponents is first determined (the G vector). The H matrix (or bridgematrix) translates the final demands by GNP component into final demandsby the industrial sector (the F vector). The technology matrix (the Amatrix) then translates the F vector into gross output by industry (the Xvector). Value added is determined by fractioning the gross output intointermediate and primary inputs.

24

TABLE A-3. SIMPLE INPUT-OUTPUT TABLE IN COEFFICIENT FORM

NetAgriculture Steel Oil Consumption Exports Other

Agriculture

Steel

Oil

Value Added a/

0.33

.17

.17

.33

0

.50

.25

.25

0

.25

.10

.65

0.89

—

.11

0.15

.77

.08

1.00

— -

—

GNP Components — 1.00 1.00 1.00

Gross Output 1.00 1.00 1.00

a/ Labor and capital inputs.

Figure A-2 shows the direction of solving the input-output relation-ship, translating final demands into industry output requirements. Thedifficulty with solving the model using this standard procedure to study theimpact of an oil production cutback is that restrictions will be applied toindustrial outputs. This means that, rather than using the composition offinal demand to determine the composition of industrial output, industrialoutputs are predetermined to some degree by the RPA petroleum allocationregulations. Defining XE to be the vector of constrained gross outputsinduced by the oil production cutback, Figure A-2 must then be solved"backwards." Thus, a new set of final demands must be derived, given XE.Although the industrial deliveries to final demand can be determined in arather straightforward manner, given the technology, a difficulty imple-menting this step arises because no unique relation exists between theindustrial deliveries vector (F) and the vector of GNP components (G).

There are several ways to circumvent the indeterminancy problem ofdirectly translating the industrial deliveries vector into the final demandcomponents. One way, used in several earlier studies of oil productioncutbacks, is to reduce final demand through some ad hoc procedure until theoutput constraints are satisfied. An alternative methodology, however, wasdeveloped for this study. The basis for the methodology can be summarizedby examining Figure A-3, which shows the direction of price conversion inthe model. Prices are built up from both unit labor and capital costs, which

25

Figure A-1.Basic Input-Output Structure

Agric. Steel Oil

Agric.

Steel

IndustrialDeliveries to

Final Demand

Consumption,Net Exports, and

Other Final Demands

ValueAdded

GrossOutputs

VAGNPComponents

Figure A-2.Direction of Solution of Input-Output Relationships

VA

X, XE

26

Figure A-3.Direction of Solution of Price Relationships

w nh P0 r1 ie cs ea se

Primary Prices

Implicit Final DemandDeflators

Unit Labor Costsand Unit Capital Costs

in turn determine gross output prices. The gross output prices are then usedto determine the implicit final demand deflators. The direction of priceconversion in the model parallels the direction of the oil production cutbacksolution—from the output side, back to final demand.

Price adjustment because of excess demand can be treated in twoways. The methodology can best be explained by writing out the basicrelationship for determining gross output prices:

i = 1, ,63k = 1, ,63k f i

27

where

w thP . = output wholesale price index for the i sector,

P. = value added deflator of the i sector,

2 « a, . P. = column weighted sum of the input wholesale prices, and

B. = the ratio of value added to the gross output for the i sector.

This relation states that the weighted sum of the primary input pricefor an industrial sector plus the material input prices yields the gross outputprice for a sector. This relation is modified in the production cutbackperiod to account for excess demand for the sector's output. The revisedspecification reads:

* = B.P7a+ Ta, . Pw + Ji 1 1 L ki k

where the additional term J represents the adjustment to prices because ofthe reduction in deliveries to final demand. J is determined by the fractionof the percentage shortfall that is passed into gross output prices.

Alternatively, the J term can be suppressed. The argument is thatprice controls permit price increases only sufficient to compensate forincreased costs of production, but a mechanism is required to adjust theallocation of final demand. The methodological approach is to bypass themodel's normal price channels and transmit the allocational information tofinal demands by way of "shadow prices"—that is, although the aggregateprice level is not affected, relative prices are. The final demands (FDj) areadjusted by the relation:

* e3FD. = FD. fP (embargo) 1

3 3 [P (preembargo)J

where FDj*, is the final demand for category j after adjustment for changesin relative prices because of the production cutback. The ej term is anelasticity parameter. These terms are assumed for each component of finaldem and.

This description has greatly simplified the price-income effects, butthe main thrust of the argument is that the proposed methodology system-atically solves for a feasible final demand set. The RPA allocation plandetermines XE, the constrained set of gross outputs. Given XE, a vector of

28

industrial deliveries to final demand (FE) can be determined. \J Using aninitial estimate for final demand (G), a corresponding required deliveriesvector (F) can be obtained. By comparing the two vectors—FE and F—on asec tor-by-sec tor basis, an excess demand vector is derived. This is used torevise the gross output prices for each industrial sector. In turn, the revisedgross output prices adjust final demand prices. For each final demandcategory, then, a revised estimate is obtained by comparing the constrainedand unconstrained final demand prices. This methodology permits either thepassthrough of excess demand inflationary shocks or the suppression of thoseshocks by some form of price controls.

Model Assumptions

Scenarios were designed with varying levels of oil supply interrup-tions and policy responses and applied to the model. The assumptions madefor the simulations used in the analysis are summarized below.

Base Case GNP Growth. The GNP projection is a real annual growthrate averaging 2.7 percent through 1987. The 1984 GNP is projected to be$4.1 trillion in current dollars.

Oil Prices. Decontrol of domestic oil prices by 1982 is assumed.Imported oil prices are projected to increase at an average annual compoundrate of 11 percent between 1980 and 1984 and 8.5 percent between 1984 and1987. The price of imported oil during an interruption is calculated by usinga price elasticity with respect to demand of approximately -0.1. After theinterruption, the price level is assumed to drop to a level 30 percent abovethe price at the start of the interruption.

Demand. Total U.S. petroleum consumption is projected to remainrelatively stable between 1980 and 1987, at about 18.3 million barrels perday. Domestic production is projected to decline from 8.5 million barrelsper day in 1980 to 7.4 million barrels per day in 1987.

Conservation. The degree to which various industries can absorbsupply cutbacks without reductions in output differs. In this analysis theaverage assumed input cutback that could be absorbed is about 7 percent.

\J See discussion of Figures A-l and A-2 for definitions of XE and G.

29

Allocation Policies. The allocation rules used to produce the resultspresented in the text are intended to minimize GNP losses. Supplies aredistributed to demands depending on the degree to which output is affectedby input constraints. A second set of allocation rules was incorporated tomodel those provided by the Energy Policy and Conservation Act (EPCA) of1975. These rules allocate supplies efficiently only after the demands ofpublic health, safety, farms, transportation, and utilities are met.

APPENDIX B. FOUR RESERVE COMPLETION CASES

Four cases were developed to explore the significance of different oilacquisition and storage construction schedules for the Strategic PetroleumReserve (SPR). Case One reflects the Department of Energy (DOE)schedule as defined in the fiscal year 1981 budget request made in January1980. Case Two assumes the highest rate of fill possible of existing storagecapacity. Case Three is an accelerated case. In addition to the high fillrate assumed for Case Two, expansion of storage capacity is increased overthe current DOE schedule. Case Four is based on linking the reserveschedule to the production levels of federally owned oil reserves.

Case One

Case One is the Department pf Energy plan as described in theJanuary version of the fiscal year 1981 budget request to the Congress. Thisplan calls for the resumption of oil acquisition in June 1980 at a rate of100,000 barrels per day (3 million barrels per month). The rate ofacquisition is to remain at this level until the second quarter of fiscal year1981 (beginning of calendar year 198?) when the rate would increase to250,000 barrels per day (7.5 million barrels per month). Construction ofadditional storage of 290 million barrels is scheduled to continue throughfiscal year 1981. This additional storage is the result of expansion ofexisting sites. No decision regarding the next expansion has been made.The budget request, however, does include planning funds for furtherexpansion, including regional storage (24 million barrels). Storage of 728million barrels is projected by 1989, assuming the completion of anadditional 190 million barrels of capacity (see Table B-l). \J

The President's revised budget, as announced in March 1980, calls forthe delay of oil purchases from June 1980 until June 1981. No changesin the development schedule were planned. This delay in oil purchasewould increase the cost of completing the reserve by approximately$0.5 billion.

(31)

32

TABLE B-l. CASE ONE—DOE SCHEDULE (By calendar year, in millions ofbarrels)

1980198119821983198419851986198719881989

StorageCapacity Available

250250308385406461562617677728

OilAcquisition

203690907755101556051

TotalOil Stored

113149239329406461562617677726

Case Two

Case Two assumes the highest rate of oil acquisition to fill existingcapacity, or approximately 400,000 barrels per day (12 million barrels permonth). This rate would fill the existing sites within 13 months (by July1981 assuming that acquisition begins in June 1980). The Department'sJanuary budget plan is followed over the remaining period (see Table B-2).

TABLE B-2. CASE TWO—ACCELERATED OIL ACQUISITION (By calendaryear, in millions of barrels)

Storage Oil TotalCapacity Available Acquisition Oil Stored

1980 248 72 1641981 250 86 2481982 308 58 3081983 385 77 3851984 406 21 4061985 461 55 4611986 562 101 5621987 617 55 6171988 677 60 6771989 728 51 728

33

Case Three

Case Three is an accelerated scenario. The highest rate of fillpossible is assumed for existing capacity, as in Case Two, along with anacceleration in the expansion of storage capacity. This could be accom-plished by using the "turnkey" approach originally planned by DOE. Theturnkey approach called for the lease of storage capacity developed byprivate firms. The Department has not proceeded with this approach fortwo reasons. First, because of the uncertainty of oil availability, the long-term leases required were not thought advisable. Second, initial contractbids were higher than DOE anticipated and significantly higher thanprojected costs of government facility expansion. While resulting in highercosts for storage capacity, this case would result in the storage of 728millions of barrels by 1986, compared to 562 million in Cases One and Twoand 482 million in Case Four (see Table B-3)

TABLE B-3. CASE THREE—ACCELERATED STORAGE DEVELOPMENTAND OIL ACQUSITTON (By calendar year, in millions ofbarrels)

1980198119821983198419851986198719881989

StorageCapacity Available

248250408585706861

1,0001,0001,0001,000

OilAcquisition

7284

160177121155141

— — —

TotalOil Stored

164248408585706861

1,0001,0001,0001,000

Case Four

The fourth case assumes the slowest fill rate, matching the rate ofoil acquisition with the rate of production of the Naval Petroleum Reserve(NPR), thus maintaining the NPR as a national reserve. The productionlevel is projected to be 165,000 barrels per day, or 5 million barrels per

34

month (see Tabel B-4). This scenario could be accomplished by using theNPR oil for the SPR or by requiring SPR purchases to match the NPRproduction level.

TABLE B-4. CASE FOUR—RELIANCE ON FEDERALLY OWNED OIL (Bycalendar year,'in millions of barrels)

198019811982198319841985198619871988198919901991

StorageCapacity Available

250250308385406461562617677728728728

OilAcquisition

306060606060606060606060

TotalOil Stored

122182242302362422482542602662722728

. Basic Cost Assumptions

Development costs for storage averaged about $3 per barrel for thefirst 248 million barrels and are expected to reach about $5 per barrel forthe next 290 million barrels by the expansion of existing sites. The thirdincrement of capacity (250 million barrels) is projected to cost up to $9 perbarrel. These estimates are based on the proposed storage costs made bythe private sector during preliminary turnkey negotiations. For Case Three,an average cost of about $20 per barrel was assumed for all storage beyondthe first 538 million barrels. This assumption was made to reflect thepremium that might be required to accelerate development.

Oil acquisition costs are based on CBO's projections of oil prices(refiner acquisition costs) over the relevant period. Although the SPR paysdomestic refiner acquisition prices, it also must pay a higher transportationcost because of required adherence to the cargo preference law and pipelinecosts. This projection is based on an annual inflation rate of 8 percent andan annual real price increase of 2 percent for oil.

35

Estimated Reserve Costs by Case. The January DOE plan (CaseOne), calls for 300 million barrels of capacity in 1982, 560 million barrels by1987, and a 728 million barrel reserve by 1990. Based on CBO assumptions,this plan results in a total project cost of $42.7 billion.

Case Two—filling existing capacity as quickly as possible—wouldsave about $1 billion of the $43 billion that would be spent for a 728 millionreserve under the DOE schedule (see Table B-5). 2J This would result in areserve of 248 million barrels by July of 1981.

TABLE B-5. SUMMARY OF COSTS FOR A 728 MILLION BARRELRESERVE (In billions of dollars) a/

StorageDevelopment

Oil Acquisition

Total

CaseOne

2.8

39.942.7

CaseTwo

2.8

38.7

41.5

CaseThree

6.3

31.7

38.0

CaseFour

2.8

42.6

45.4

a/ This excludes expenditures through fiscal year 1980, which total$0.9 billion for development and $1.3 billion for oil acquisition in previousyears.

The accelerated scenario (Case Three) shows that, for a 728 millionbarrel reserve, about $4.7 billion could be saved from the DOE schedule,despite the higher development costs necessary to maintain sufficientstorage capacity to meet an aggressive fill schedule. A total of 728 millionbarrels would be stored by 1986 under this scenario, compared to same levelin 1989 under DOE's current plan. A 1 billion barrel reserve could becompleted by 1987 under the accelerated schedule at a total cost ofapproximately $62 billion. A slower development schedule would result ineven higher costs. Whether such a schedule could be achieved would dependon both the availability of oil and how quickly storage capacity could becompleted.

2J The savings would be as much as $1.5 billion when compared to theadditional purchase delay called for in the President's March update tothe fiscal year 1981 budget request.

36

Case Four—linking oil fill to production from federally owned oilreserves—is the slowest of the cases and would cost the most. The728 million barrel reserve would not be completed until 1991, at a cost ofmore than $45 billion—almost $3 billion more than DOE's plan.

Estimated Reserve Costs by Size. The costs of a number of reservesizes were estimated assuming the varying schedules of the four cases. Tocomplete a 248 million barrel reserve would cost between $6 and $7 billionin addition to about $770 million for development and $1.3 billion for oilspent prior to fiscal year 1980, depending on which case is assumed.

The cost of a 538 million barrel reserve is estimated to be between$25 billion and $27 billion in addition to the funds spent through fiscal year1979, the lower cost resulting from an accelerated schedule. The costs tocomplete a 728 million barrel reserve also depend on the schedule assumed.As indicated in the discussion of the various cases, costs for a reserve ofthis size range from approximately $41 billion to $45 billion, the lower costagain resulting from an accelerated schedule. The final goal of 1 billionbarrels will cost at least $62 billion based on these estimates. Theaccelerated schedule reaches this goal by 1986; no other schedule reaches abillion barrels until after 1990. Later achievement of this goal will likelycost more than the accelerated case.

o